DCAudioDIY.com

For those wondering about digital crossovers and their differences to passive crossovers, here is a white paper from me to start discussions.

This is one discussion point. Another is that the digital crossover can do long time delays to time align multi-cabinet speakers system like the line arrays I am working on.

I agree with Roscoe that the DC resistance of the VC largely makes damping factor a myth unless the amp has significantly higher output impedance than the the coil resistance. If you have servo drive it is an entirely different issue because the servo can drive the amp output to greater voltage levels than would be dictated based on a fixed gain to put enough current through the voice coil to make the cone behave.

However after that, digital, or for that matter any active crossovers have huge advantages over passive designs. Ask David Berning the size, cost, and complexity of the passive crossover he used for his sub. Even after that he had to use an active EQ to get the balance correct. Passive crossovers among other things are susceptible to non linear impedance in the driver in the crossover region. If the impedance were non linear (gosh, I could hardly imagine that ) the slope of the crossover would also be non linear. With an active crossover, any low output impedance amp would not care about that and the point and slope would be as desired. Eliminating all the non linearities and other problems with crossover resistors, capacitors and inductors which are absent in an active configuration greatly lowers signal degredation. DSP also allows other controls not specifically related to the prime crossover function such as time delays, phase shifts, and often parametric EQ and dynamic EQ. I have not built a passive crossover in perhaps 40 years. They are just too difficult, expensive, and limiting. There are some very good passive crossovers out there but they are hugely more expensive than an active solution even given the cost of extra amps. BTW another advantage of active solutions is the reduction in IM distortion in the amps because they handle narrower bandwidths.

When using active crossovers and directly powering each driver care must be used to prevent DC in the mid and high drivers. With passive crossovers capacitors block DC in those regions. Woofers are still at risk in either case.

Tom

Tom,
Agree. Just also showing how the simple case of speaker cable resistance will influence the output impedance of the amp even with fairly decent size 14awg cable with a typical length. I see power amps rated with a DF of 1100 and wonder if most folks really understands what that means.

The simple one IC DC offset protection circuits available for DC coupled amps help eliminate the problem if there is a fault in the amp and get DC on the output. All the amp modules so far have been well below 10mV of DC offset and no problems.

Rich

_________________
Walt

The Nelson Pass crossover is an analog active crossover with similar issues to passive speaker level crossovers. This talk is about digital active crossovers. Very different animals.

The myth that today's active devices are inferior to passive devices and combined with effects passive devices have at speaker signal levels is just that, and old wives tale. Although there are always different paths to any solution, the passive path is full of more ruts and cliffs to fall off of than the active route. When you start with a philosophy that is fraught with problems you have to work much harder and spend a lot more money to overcome them. It can be done as is evidenced by high quality vinyl solutions but you really pay the price in money, time, and reliability to achieve at best an equivalent performance. Doing stuff the old way can be very satisfying and a valid goal in itself but will inevitably put major constraints on your resources.

Just saying

Tom

_________________
Walt

And one visit here might do the same.

Now back to digital crossover discussions.

Did a very interesting listening trial with the DSP crossover I use on my speakers.

Created a L-R 8 Pole, 2-way crossover and summed the outputs. The theory is that a L-R will sum perfectly when the high pass and low pass filters are used, so summing them should give flat response.

Ran the crossover as a HiRez DAC and listened with my planar magnetic headphones. Cannot tell the difference between that and just a straight connection, even listening through 16 poles of digital filters. Sounds excellent to me.

Makes me happy!

I listened through headphones to a recording that Rich passed back and forth through his all pass digital filter and it sounded great.